Composite Construction in Steel and Concrete IV

In September 2000, a new Highway Bridge Design Code, based on limit states design principles, will be re-established for all of Canada. The highlights of and rationale for the design provisions for composite beams and girders in this code are presented. These provisions have been greatly simplified and rationalized as compared to previous codes. A unified approach is proposed for evaluating the factored moment resistance of Class 1, 2 and 3 steel sections (plastic, compact and non-compact sections, respectively), and stiffened plate girders. Composite beams and girders with steel sections, symmetrical or unsymmetrical with respect to the flexural axis, are treated similarly. An equivalent plastic stress distribution technique, which is achieved by neglecting portions of the web, is used for the evaluation of the factored moment resistance of the most slender sections in positive moment regions. This method was calibrated against a more precise but more cumbersome elasto-plastic stress distribution technique, as well as against existing code recommendations. The factored moment resistance of Class 3 (non-compact) sections and stiffened plate girders in negative moment regions is the only resistance based on a linear stress distribution at first appearance of yield in the steel section.

A study of the causes of transverse cracking in concrete slabs of continuous composite bridges has been carried out in order to understand better the most important parameters with respect to the origin of tensile stresses in the slab. It was demonstrated that for typical bridges the most critical tensile stresses in the slab are induced during construction The results of in-situ and laboratory measurements have shown the important influence of concrete hydration and the pouring sequence on these tensile stresses. Indications for methods to limit tensile stress have been given as a function of the span length and of their short and medium-term efficiency.

Current developments in the field of steel-concrete composite bridges focus largely on the durability and in-service performance of the concrete deck slab. Indeed, one of the major concerns of designers is to attempt to limit slab cracking, especially full-depth transverse cracking. Amongst the most frequently suggested measures applicable to steel-concrete composite bridges, longitudinal prestressing remains a feasible solution for composite structures where transverse cracking is to be limited or eradicated altogether. The aim of this investigation is to review current prestressing techniques, to investigate the behaviour of the deck slab over a period of time, to establish the most suitable parameters to limit tensile stresses in the slab and to develop a simplified approach for the calculation of prestressing losses.

This paper presents results from repeated loading tests on composite beams. Repeated loading causes an increase in the deflection of composite beams due to incremental slip in the shear connection. An analytical model to calculate this is presented, and comparisons are made between the analytical model and the test results.

The influence of realistic stress range histories on the fatigue behavior of headed studs in composite bridge constructions is investigated. The validity of Miner's rule and the influence of profiled sheeting is checked by experimental testing. A simple modified Miner's rule for massive concrete slabs and a basis for the design with profiled sheeting are derived.

The short-term elastic and long-term time-dependent behavior of extradosed bridges with composite girders are studied. Analysis is carried out with a FEM step-by-step method, which was developed originally at Yamaguchi University. It is confirmed that large changes of the deflection and the longitudinal stress distribution occur and the stresses in the concrete slab increase in compression for a long period.

In continuous girder bridges a large negative moment occurs in regions over the intermediate piers. In order to prevent cracks in the concrete slab, the authors have developed special connectors, which are flexible enough to transmit little shearing force, enduring many cycles of deformation. They have already been applied to actual railway composite girder bridges without troubles. It was analyzed how the length of zone provided with the flexible connectors influences the characteristics of structure.

Test results are compared with theoretical predictions for a series of composite beams made with a pretensioned high-strength concrete tension flange, a corrugated steel web, and a reinforced concrete compression flange. The results from five tests indicate that this type of beam can be designed rationally. No unexpected problems were found.

A procedure is presented for checking that composite beams and frames designed according to the principles of semi-continuous construction possess sufficient ductility to ensure that the presumed pattern of moments at the ultimate limit state design condition is achieved. The procedure provides for this in terms of the available and required inelastic rotations at each notional plastic hinge in the system as devised by detailed evaluations of numerous previous studies of the subject area.

In Eurocode 4-1 the design bending resistance for distorsional lateral buckling is based on a reduction of the plastic design bending moment. The reduction factor depends on the elastic critical bending moment M_cr. It is shown that the elastic critical bending moment of composite beams can be determined with help of the analogy between the differential equilibrium equations of the compression member on elastic foundation and the lateral torsional buckling problem. Based on this analogy a new method for the determination of elastic critical bending moment is derived and compared with the method given in Eurocode 4-1.

In special cases partially encased composite beams are used without concrete slabs. In such cases the beam may fail by lateral torsional buckling. No information is available whether or not it is allowed to take into account the torsional stiffness of the concrete and if so to what amount. 22 full scale tests were carried out on beams loaded by a point load. The test results were evaluated using different assumptions which lead to a new design proposal.

The design of composite beams using rational partial shear connection theory is well established and has been included in design codes such as Eurocode 4 (1) and the more recent Australian Standard AS 2327.1-1996 (2). However, the design of composite slabs has previously been based on empirical methods. Australian steel decking profiles exhibit strong and ductile shear connection to the concrete slab and can be designed using partial shear connection theory in a similar manner to beams, thereby providing a rational design approach to flexural members.

In order to investigate the influence of reinforcement on the development of cracks and the behaviour of composite slabs as part of composite beams in hogging bending areas, experimental investigations and parameter studies by calculation have been carried out at the University of Kaiserslautem. These studies are described in (1, 2, 3, 4). In this paper the main results of the studies are described. The results in details show. 1) The development of cracking in slabs as part of composite beams is decisively influenced by the transverse reinforcement. Proposals to determine the minimum reinforcement with reference to the service load and to calculate the crack width in dependance on the transverse reinforcement must be determined. 2) The experimental investigations with composite beams showed, that in view of the load-bearing capacity different kinds of minimum reinforcement exist: one to reach the plastic moment and one to make the application of the plastic hinge theory possible. Conditions for composite beams to reach both kinds of minimum reinforcement are determined.

This paper is a summary of an American Institute of Steel Construction report on the development of low floor-to-floor height systems using structural steel. All systems include composite action between structural steel and concrete. A list of evaluation items is presented and discussed. Each system is described and rated for potential growth in both the residential and commercial market.

Shallow floors in which the main part of the steel section resides within the concrete slab depth have increased in popularity in many European countries, Scandinavia included, and there are a variety of components available for these, depending on what part of Europe is concerned. Independent of the type of steel member employed, three main types of structural system may be recognized: (1) a steel member supporting hollow core decking units, (2) a steel member supporting composite slabs with deep or shallow profiled sheeting and (3) a steel member supporting concrete composite slabs with prefabricated concrete planks which are reinforced or prestressed. This paper discusses the experimental and theoretical behavior of the three classes as observed in the tests and as analyzed by the non-linear finite element method. Design resistance models based on the idea of partial strength shear connection behavior for the ultimate limit state and stiffness models for the serviceability limit state are proposed and compared with experimental and theoretical behavior data.

This paper presents the results of a numerical analysis modelling the behaviour of the continuous stem girder system, which comprises two end cantilever stems and an interspan composite element. The analysis includes two connection models; the stem to interspan beam connection and the shear connection of the composite central beam. The numerical model is extended to applications for slim floor systems which utilise the new British Steel range of asymmetric sections which have unequal flange widths. Only linear elastic behaviour has been investigated. However the test results are sufficiently encouraging to warrant the study being extended to include post elastic behaviour and additional loading configurations.

An experimental investigation on a novel stub-girder system, incorporating the Slimdek concept, is discussed. The test results from a full-scale 12.6 m span prototype composite girder, which was load tested to a mid-span deflexion of span/44 (i.e., 285 mm) at the University of Cambridge Engineering Department, show that this structural solution has great potential for spans in the 12 to 20 m range; where a high degree of service integration is often demanded. Particular advantages inherent within this structural concept are that: explicit mechanical shear connexion between the girder and the concrete slab is eliminated; unpropped construction may be adopted; and, due to the long-spanning capabilities of the deep composite slab, secondary beams may be eliminated.

Corus UK Ltd has developed a new system for use in double skin composite panels (Bi-Steel). Analytical and a numerical models have been developed to gain insight into the physical behaviour of these plates. In this paper both models as well as design rules for the verification of sandwich elements are presented.

Thel performance of steel-concrete composite structures depends significantly upon the bond characteristics of connection between concrete and steel plates. A new type of shear connectors was developed using perforated steel plates, commonly known as `'perforbond'. An experimental investigation was conducted to study the shear strength characteristics of such connections, considering several test parameters. This paper describes the results of the experimental investigation. Further, based on the experimental results, prediction equations are proposed to calculate the shear strength of perforated steel plates and effects of confinement due to penetrating steel.

The finite slip capacity of mechanical shear connectors can have a significant effect on the moment capacity and behaviour of a composite beam. This issue becomes especially important when partial shear connection is adopted, since fracture of the shear connectors by excessive slip can cause premature failure of the composite beam. This paper outlines a generalised mathematical model for continuous composite beams that can be adopted in identifying the moment that causes fracture of the shear connectors. Calibration against existing experimental results is outlined, and a parametric study considering beam length is also discussed. It has also been identified that the model needs to be extended to allow for concrete cracking and material non-linearity to allow a proper treatment of the shear connector capacity.

Continuous composite beams are affected by negative bending moments at the internal supports; thus tensile stresses arise in the concrete slab. Cracks in the concrete slab and slippage between concrete and steel component make the structural response very complex in the hogging moment regions. Thus behaviour results from many non linear phenomena involving the reinforced concrete and the steel components, and also very localised effects due to bond of the reinforcing bars and shear connection. Recent developments in the field of modelling of such beams are presented and a theoretical model able to take account of partial interaction and tension stiffening effect is discussed.

Six full-scale steel-concrete composite substructures with headed stud shear connectors and full and partial shear connection have been tested by the authors. In this paper, the corresponding inelastic responses to different sets of variable and random reversed displacements and the importance of cyclic loading on the strength softening and the ductility deterioration of members are scrutinized. Finite element simulations are established to reproduce composite beam behavior and to evaluate different modeling assumptions and local effects. Analytical issues related to the effective width of concrete deck, to the force-slip law of shear connectors and to the beam shear connection degree are investigated.

This study addresses the cyclic, nonlinear analysis of frames composed of steel columns, steel-concrete composite girders and partially-restrained composite (PRC) connections. The effect of the slab around the column is modeled by partial-composite beam finite elements, component-based connection models, and slab load transfer elements. The key characteristics of this approach are outlined via an example analysis of a two-bay experimental test.

The behavior of and design methods for longspan composite joists are presented. Both strength and serviceability limit states are considered. Studies of composite joists and joist girders conducted at Virginia Tech are highlighted. Design provisions available to the structural engineer are identified and reviewed. Several buildings constructed using composite joists are discussed.

Composite floor trusses Supporting composite wide rib floor deck and a concrete cover slab were used successfully on a 51 story high-rise building, "Le 1000 de La Gauchetière" located in Montreal. The steel structural proposal for the building was selected in a competitive design build tender over several alternate concrete structural solutions. The economy of the floor system was one of the contributing factors leading to the selection of the steel solution. The theory and design of composite trusses is reviewed, and equally important, the fabrication aspects and selection of individual components that permitted the composite truss scheme to be the most economical and viable floor system is discussed.

Brittle rib shearing failure comprises horizontal splits that form in the concrete between the tops of the steel sheeting ribs in pans where there are shear connectors, while the failure surface locally avoids the studs by passing over their heads. It has been reported in the literature to have often occurred in push-out and beam tests. It is now mandatory in Australian Standard AS 2327, Part 1 to reinforce composite edge beams against rib shearing failure. A patented reinforcing component has been developed for this purpose.

The aim of this paper is to present a design method based on small scale tests and a finite element analysis called 3P PCSM (three parameters partial connection strength method) for composite slabs with shallow sheeting. A short review of the current design methods in Eurocode 4 (1) are given just to accentuate a new contribution. Distribution of forces within composite slabs for various loading is discussed. A well established definition of the shear length is redefined in the context of the new findings, in Veljkovic (2) and for implementation in the new design method. Finally, an example is given where a design according to 3P PCSM is compared with two methods given in Eurocode 4, the m-k method and the partial connection strength method.

Composite slabs are simple and economical construction elements. However, the verifications that are required for their design (structural safety, serviceability) are long and complicated. Current design methods found in standards and guidelines rely on the results of costly and time-consuming large-scale laboratory tests. A new design approach for the prediction of composite slab behaviour is proposed. The new approach combines results from standard material tests and so-called small-scale tests with a simple calculation model, the New Simplified Method, to obtain the moment-curvature relationship at the critical cross-section of a composite slab. Unlike recently proposed methods, the calculation method described herein does not rely on computer dependent numerical simulations. Results obtained from using the new design approach have been verified through comparisons to large-scale tests. For greater acceptance and uniformity of the results obtained from the New Simplified Method, the development of a standardised small-scale laboratory test to measure the longitudinal shear versus slip behaviour at the steel-concrete interface is discussed.

A new model is presented describing the physical behaviour of shear connections in composite slabs. The model recognizes the improved shear resistance over the supports of shear connections in composite slabs. The application of the new model is related to the Partial Shear Connection method as available in the ENV 1994-1-1 (Eurocode 4). The model should be considered as a first step towards a design model based on obtained insight into the physical behaviour of connections in composite slabs.

The influence of the "accordion-effect" on the determination of the longitudinal shear resistance and the bending resistance for deep profiled sheeting is discussed and some results for the SD225 profile are presented Besides this, first results of a slab-test with a line load parallel to the ribs are shown and some simple rules for a safe design of the slab under line loads are derived.

Two point loading tests of twenty three simply supported open-sandwich beams which consist of concrete and embossed or plain steel plate with headed studs, were carried out. Two types of embossed plates were used, namely checkered and ribbed types. Ultimate loading capacities and failure modes were examined in relation to the types of steel plates, stud spacing and shear span ratios. Methods to predict the shear bond strength of an open-sandwich beam are discussed.

Perforated shear connector in several modifications is presented. The connector was tested in laboratories of CTU Prague for static shear load capacity. 61 push tests of connectors with 32 mm openings and with 60 mm openings, both in normal and lightweight concrete and with different transverse reinforcement, were carried out. The tests also include behavior of connectors with modified "high" position of 32 mm openings and of two connectors in parallel arrangement. After statistical evaluation of the experimental results the authors developed reasonable shear resistance formulas for all proposed arrangements and suitable for structural design.

This paper presents slip behavior of perfobond rib shear connectors on push-out and girder bending tests and their treatment in FEM analysis using plates and combination elements. The number of holes and thickness of ribs are varied in twelve push-out test specimens. The combination elements in FEM between concrete and the perfobond ribs are modeled by spring constants to express the applied load-slip relations.

Statistical analyses of test data show that: (1) for stud connectors in troughs in profiled sheeting, existing design rules omit important parameters; and (2) for studs in fatigue, the wide range of proposed S-N lines arises from differences between methods of analysis of the test data. (3) Finite-element analyses lead to rules for the layout of shear connectors near concentrations of longitudinal shear in bridge beams.

Composite beams formed by connecting the concrete slabs to the supporting steel beams have been in use for many years. Much of the research into this form of construction has concentrated on the more traditional reinforced concrete and metal deck construction. In composite beam design, the strength of the shear connector is of great importance since longitudinal shear forces must be transferred across the steel-concrete interface by the mechanical action of shear connectors. The shear strength and stiffness of the connection is not only dependent on the strength of the connector itself, but also on the resistance of the concrete slab to longitudinal cracking caused by the high concentration of shear force at each connector. Present knowledge of this behaviour is limited to shear connectors in solid reinforced concrete slabs and concrete slabs with profiled sheeting, little information is given for the shear connector capacity on composite steel beam with precast hollow core slabs. A new horizontal push off test is proposed to determine the shear capacity of the connector on the composite beams with precast concrete hollow cored floor slabs. The results showed the new test is compatible with all type of floor and shear connectors, and can replace the existing vertical push off tests.

This paper briefly summarizes the results of an experimental program which involved the testing of 52 push-out specimens and four full size composite beams with wide ribbed parallel metal deck. A regression analysis of the test data of the push-out specimens yielded a new equation, which provides much better correlation to test results than those obtained using the current provisions. The experimentally determined ultimate flexural capacity of the full size composite beams agreed reasonably well with the predicted values based on the proposed equation.

Novel shear connectors are presented as an alternative to welded headed stud connectors for composite beams. They are attached to the beam by means of powder-actuated fasteners. The design of the connectors allows the use of automatic tools to improve installation efficiency. The aim of the development is to achieve high shear capacity per fastener with sufficient ductility to allow plastic design. Various design solutions were tested with pushout tests according to Eurocode 4 (1). The objective of the pushout tests was to investigate the resistance and the ductility of nailed rib shear connectors. One of the designs, the rib connector, consists of two folded angles made from a thin steel sheet. One leg is perforated and embedded in the concrete; the other leg is fastened to the beam by means of powder-actuated fasteners. Based on the test results rib connectors can now be designed to achieve plastic shear behavior and develop characteristic loads of more than 18 kN per nail with a shank diameter of only 4.5 mm.

Innovative forms of composite cross sections lead to an unusual positioning of the headed studs horizontally in the thin concrete slab. By experimental and numerical methods the behaviour of lying studs with longitudinal shear force has been investigated. The results of about 50 push-out tests and numerous calculations show that the failure characterised by splitting of the thin slab is influenced by different parameters compared to vertically positioned studs. Based on the investigations design rules for lying studs, including a formula for the design resistance, are presented.

Tests to establish a Manson-Coffin relationship for headed studs as shear connectors in composite beams are reported. The joint shear displacement was controlled and the resulting force amplitudes measured. Strain gages monitored each stud and allowed for a separate evaluation of stud failure. Although a considerable variation of stud loading was observed, failure occurred over a rather narrow range of cycles. Like any Manson-Coffin relationship, the one established here for studs covers the full range of fatigue (low cycle and high cycle fatigue). This is a very economical way of determining fatigue resistance. In the future, Manson-Coffin relationships for headed studs should be developed for different amplitude collectives, stud diameters and concretes.

The behavior and strength of welded shear studs are subjects of ongoing study. A description of the formulation of the American Institute of Steel Construction specification provisions for the strength of shear studs is given. The behavior of shear studs as influenced by friction, position of studs in steel deck and stud properties is described. A new model for strength determination is introduced and compared with data from tests conducted by the authors as well as others.

Three series of pull-push specimens with 16 mm and 22 mm diameter headed stud shear connectors were built and tested as part of a general investigation on seismic design of steel-concrete composite beams with full and partial shear connection. To assess the shear connector performances from a seismic and damage standpoint, pull-push specimens have been exposed to series of variable, random and constant reversed slips. Main results are commented upon and evaluated in terms of yielding and maximum shear strength capacity as well as ultimate slip ductility. A comparison between experimental and prediction strengths derived by relevant design code provisions provides an estimate of their accuracy. Finally, some low-cycle fatigue models are investigated to establish damage limit domains and fatigue life relationships for headed studs.

This paper presents the final results of an experimental and analytical study designed to develop a new shear connector for composite steel-concrete structures. Innovative characteristic lies in the design features of the connector's base, which presents a special folded shape. A modified push-out test has been employed to assess the shear strength as well as the load-slip behaviour under cyclic loading. A remarkable strength and ductile behaviour have been observed. In this way, an optimised and economic shear connector was obtained.

Structural systems that combine steel and reinforced concrete provide increased stiffness and resistance at minimal cost, but stress transfer between the two materials is needed. This paper addresses these stress transfer requirements. The composite structural systems used for resisting lateral loads are reviewed, and the stress transfer demands for different systems are described. These demands are often quite large and localized. The stress and force transfer capacities are then evaluated, and design strategies for concrete filled tube (CFT) and encased steel section (steel reinforced concrete or SRC) are presented. The work shows that mechanical shear transfer devices can be avoided with many SRC applications, but significant transfer problems exist for braced frames with CFT columns. An ongoing research study into the requirements for braced frames with CFT columns is briefly described.

This paper describes the research program that has been performed to develop and validate design equations for the axial capacity of partially encased composite columns made with three plate built-up steel shapes. The width-to-thickness ratio of the plate elements in these columns exceeds the limit for non-compact section and local buckling had to be examined for bare steel columns carrying construction loads as well as for composite sections. Potentially adverse effects of the loading sequence in actual columns and the confinement and creep of the concrete on the column axial load capacity were also examined in the study. Design equations are proposed for the bare steel columns and the composite columns.

This paper presents the development of a database for composite columns, and discusses its use in calibrating new design procedures for these structural elements. Current design procedures are shown to provide both lower reliabilities than codes imply and significant scatter of the results. Seven alternatives to current provisions are proposed and their performance examined. Approaches based on the full plastic capacity of the section are shown to provide a good starting point for code provisions, and two alternative schemes for improving current provisions are proposed.

The rules of Eurocode 4 (1) apply for normal strength steel and normal strength concrete. In order to check the rules for the use of high strength steel a theoretical investigation based on Finite Element calculations had been carried out (2). Within the framework of the European research project "Use of high strength steel S460", experimental static strength investigations were carried out on 32 composite columns and 2 steel columns (for comparison purposes) using high strength fine grained steel S460. In addition to using concrete qualities C20, C45 and C70, the arrangement of mechanical shear connectors, longitudinal reinforcement as well as the transverse reinforcement (stirrups) and the loading were varied. The paper will give a survey about the test results with specific observations. The tests are evaluated according to Eurocode 4.

Partially encased columns show large ultimate loads needing only minimal space. This can further be improved by using high strength materials. In this paper we present tests on partially encased columns using high-strength steel. Special emphasis was laid on strength tests of high-tensile-steel under fire condition because the steel of the flanges would be directly exposed to high temperatures in the event of a fire.

Building code seismic design provisions for steel and composite moment frames and their implications on seismic performance are examined through comparative trial designs of three six-story buildings. Using nonlinear analyses, static and dynamic contributions to inelastic force reduction are identified and compared to code-specified assumptions. The trial design study demonstrates that static overstrength is significantly larger for space frame systems compared to perimeter frames, whereas the force reduction attributable to inelastic dynamic response is fairly constant between the frame types. The trial designs are shown to meet the minimum performance implied by building code provisions, primarily due to the large static overstrength resulting from minimum stiffness requirements imposed through seismic drift limits. The study suggests that more consistent reliability can be achieved by disaggregating the R-factor into its component static and dynamic parts and that the default values of static and dynamic inelastic response factors should be reduced from those implied by current R-factors.

The use of concrete-filled steel tubes of circular cross-section (CFST's), as composite columns in high-rise office and residential construction is now commonplace. Significant economies may be gained when the steel tube is thin-walled, with diameter to thickness ratios exceeding 100. Current design of these members is based on limiting the diameter to thickness ratio to ensure that local buckling does not take place, but the specification of such a limit does not appear to have a sound basis derived from the application of engineering mechanics. Furthermore, with an accurately-determined limit, if local buckling does indeed occur then it is necessary to be able to predict the strength of the locally-buckled tube. This paper presents a Ritz-based procedure for deriving the elastic local buckling stress for the steel component of a CFST. The local buckling stress is then used in a calibration procedure for determining both the limiting depth to thickness ratio that delineates a fully-effective and slender section, that incorporates the effects of initial geometric imperfections and residual stresses.

The behavior of concrete filled tube columns under seismic loads was studied through testing six columns which were subjected to a constant axial load in addition to a cyclic lateral load. The test parameters included the level of axial load, the diameter-to-thickness ratio of the steel tube, and the concrete strength. The test columns showed high ductility and maintained their strength up to the end of the test. The test results also indicate that the column capacity was significantly improvement due to the enhancement in the concrete strength due to the confinement provided by the steel tube.

Concrete filled steel tubes are efficient members in structural applications including bridges, buildings and piled foundations. Their design has been based on theory and tests on columns under load applied axially or at relatively small eccentricities. For other cases, their design has been extrapolated due to little research. The effects of both local buckling of thin walled steel tubes and the confined concrete in flexure have not been adequately examined. This paper focuses on numerical studies carried out to determine the significance of factors such as initial imperfections and the confined concrete. These results are being used to design an extensive experimental program that will be used to develop safe economical design rules which can be directly included in the Australian Standard for Composite Structures currently under development.

This paper considers the cross-sectional strength behaviour of concrete filled steel box columns. The types of columns considered have very thin steel walls and have been used very recently to provide economy of construction for columns in multistorey buildings. A cross-sectional analysis method is developed which employs a rigid plastic analysis procedure to ascertain the strength interaction envelope for axial force and bending moment. This rigid plastic analysis method assumes the concrete is fully crushed, the steel has fully yielded and it also incorporates the effects of local and post-local buckling of the steel in compression regions of the column using an effective width model. A parametric study is conducted to consider the effects of varying concrete compressive strength and steel plate slenderness limits. A regression analysis is then carried out in order to determine a generalized model for the strength analysis procedure. This has enabled the development of generalized design equations useful in determining strength envelopes for design. The generalized approach is calibrated against independent test results and found to be conservative in its prediction of the strengths of the tested cross-sections.

Finite element analysis is carried out for two types of concrete filled steel tubular columns. A new interface element is proposed to express the behavior of the boundary between both materials. From the results, this analysis can predict the cyclic behavior and the ultimate strength of the columns with sufficient accuracy.

The Cyclic behavior of concrete filled steel tubular columns are investigated by experimentally comparing two types, bonded and unbonded. The results showed that both types have excellent ductility. They also showed that creep would increase ductility of bonded type, however it decreases ductility of bonded type.

The authors have conducted a test on concrete filled tubular (CFT) members with sizes used in ordinary building frames to investigate the scale effect on the fracture after cyclic local buckling. The comparative study with a previous test on small-scale specimens revealed that the scale effect on the fracture might be negligible for CFT with ordinary sizes and, therefore, a formula to predict fracture, which had been derived from small-scale specimens, could be used on large specimens. The applications of the formula are demonstrated for portal frame specimens with CFT braces and for CFT beam-column specimens.

Recent developments in Japan on the research, design and construction of a framing system consisting of concrete-filled steel tube (CFT) columns and H-shaped beams are presented, introducing the work done in a five-year research project on CFT column system carried out as a part of the U.S.-Japan Cooperative Earthquake Research Program.

Steel/concrete/steel sandwich construction combines structural efficiency with robustness, but has been difficult to construct. The Bi-Steel^TM composite addresses these difficulties by using shear connectors extending from plate to plate to create semi-rigid panels. The structural behaviour of the composite is described with its relevance to design. Advantages of the system are summarised along with typical applications.

In the course of the reorganisation of the scientific- and business-parc Berlin-Adlershof, the Innovation Center of Information Technology was built in 1997/1998 on behalf of WISTA Management GmbH. The architectural design allowed for an innovative concept with regard to german conditions. The system requirements and the desire to leave the primary construction largely visible with exposed steel led to a composite construction with some special technical solutions: 1) The Slimfloor ceiling system consists of axial beams with asymmetrical cross-section and a 210mm high trapezium profile. The total construction hight is 320mm with a span of 7.2m support grid and design in F90. The shear-plate-effect of the ceiling system had to be guaranteed for high horizontal forces; the concrete cover above the beams' flange amounted to only a few centimetres. This called for shear-stress to pass across the steel girder. 2) The main columns of the building form an inverted pyramid system. Considerable stabilization forces thereby develop as well as locally concentrated load-discharges at pivot-points. These are taken over by massive steel parts which project into the columns. Through additional pneumatic concrete-filling of the tube sections a composite construction develops, which at the same time guarantees the fire protection of the connection areas. 3) The storey columns were built with square pipes, which were filled with reinforced concrete to have F90 quality (90 minutes of fire resistance). 4) Further planning requirements required unusual design-solutions in detailing, e.g. i) shear plates with a hight shift ii) suspended conference rooms iii) suspension of the lower glass-facade The building was completed and put into operation in February 1999.

Results from an extensive numerical parametric study of composite frames designed according to the principles of the Wind Moment Method are used to highlight several detailed aspects of behaviour. These confirm that the proposed approach leads to frames that satisfy all important design limitations.

A design method for braced frames comprising semi-continuous partial-strength composite joints is proposed. In this design method, the ultimate load carrying capacity is calculated considering the plastic moment resistance of the composite joint in the hogging moment region as well as a defined percentage of the plastic moment resistance of the composite beam in the sagging moment region. The defined percentage depends on the required rotation at the support and is based on the assumption of a certain available rotation of the composite joint. This assumption is assured by design recommendations for composite joints on the basis of existing experimental studies.

An experimental investigation was conducted into using 350 and 450 MPa thin-walled steel tubes in simply-supported composite beams. It was demonstrated that single welded studs or channel shear connectors can be used without affecting their shear capacity, provided the tube wall thickness is at least 4.0 mm. However, stud shear connectors must be manually welded to thin tubes to avoid bum-through. Several factors (viz. residual stresses developed in the tubes during their manufacture, and stud shear connector base flexibility) reduced the flexural stiffness of the beams. Thus, an allowance for their effect must be made in design when calculating vertical deflections in accordance with Australian Standard AS 2327.1.

A numerical model is proposed for calculating the moment-rotation curve of composite joints up to failure under monotonically increasing loading. This model is based on the component method in which new components are introduced. These new components are the steel-concrete shear connection, the concrete slab reinforcement (with or without allowance for the tension-stiffening effect) and the flexural bending of the bolted flush end-plate between the upper bolt-row and the tension flange of the steel beam. The model is favorably compared with several test results for a double-sided joint configuration under balanced loading carried out at INSA-Rennes.

In order to clarify the elasto-plastic behavior of the frame consisting of concrete-filled steel tube (CFT) columns and H-shaped steel beams under seismic loading, portal frame specimens were tested under constant vertical loads on columns and alternately repeated horizontal load. Experimental parameters were: axial load ratio (0.15, 0.3 and 0.5) and width-to-thickness ratio (21, 39 and 54). D_s factors which are the factor used in the seismic design practice in Japan, were calculated from the load-deflection curves of all specimens. It was observed that all specimens showed fairly stable hysteresis loops, and the earthquake resistant capacity of a CFT frame was better than a similar steel frame.

A CFT truss, in which the chord members comprise concrete filled tubes [CFT], may have an excellent capacity to deform. In this paper, a prediction method for the deformation capacity of the truss is proposed both for monotonic loading and cyclically repeated loading. The latter is determined based on the energy dissipation capacity of CFT chords. The predicted deformation capacity of CFT trusses showed a good agreement with test results, and the capacity is comparable to that of full web steel members.

Reinforced concrete tanks are often used to store liquids in corrosive environs. This inevitably leads to deterioration, which requires either replacement or repair and rehabilitation. When a tank is a vital component in an industrial process, downtime must be minimized and repairing it using steel plates acting composite with concrete can be a rapid and reliable method. This paper describes the rehabilitation of two deteriorated sulfur-storage tanks using composite steel liners. Two construction schemes were used: one in which the tank bottom and wails were repaired separately and another in which the entire liner was prefabricated on-site and installed as a unit. Both design and construction considerations are described. Repairs were completed in 1995 and all indications are that the rehabilitated tanks are performing as intended.

Composite steel/concrete slabs, consisting of cold-formed profiled thin-walled steel decking and concrete cast on top, are a popular flooring system in modern steel framed buildings. The fire resistance is an important performance criterion. The behaviour of composite slabs in fire is rather complex, amongst others due to the highly non-linear material behaviour at elevated temperatures. In order to determine whether the behaviour in fire meets the requirements as specified in the building regulations, simple calculation rules are available, which are based however, on limited experimental evidence, and are therefore lacking a fundamental basis. The main goal of this research was to provide this fundamental basis and along with it the establishment of new more accurate calculation rules. A significant part of the work has been carried out within the scope of an ECSC co-sponsored research project From a literature survey it became apparent that with a view to validation of numerical models, experimental evidence on the behaviour of fire exposed continuous span slabs was scarce. Full scale well-instrumented fire tests have therefore been performed on various types of composite slabs to study the thermal behaviour near internal supports and the mechanical behaviour of continuous two-span systems. The parameters investigated comprise: reinforcement ratio, load level, insulation of internal support and type of steel decking (Section 2). Finite element models have been developed and validated for the simulation of the thermal and mechanical response. The numerical models were validated on the basis of experimental evidence and appeared to be suitable for parametric studies (Section 2). A new set of simple calculation rules for the assessment of the fire resistance of composite slabs has been established. These are based on extensive parametric studies, in which the geometry of the steel decking was varied systematically and which furthermore comprised all existing types of steel deckings. Intermediate results of the parameter study were confronted with the existing calculation rules to show the improvement (Section 3).

Slim floor composite slabs are made of asymmetric steel beams with a wide lower flange supporting composite decks with profiled steel sheets or prefabricated concrete elements (mainly precast hollow core elements). The gap between steel beam and hollow core elements is filled with concrete. Models for the fire behaviour of slim floor slabs with hollow core elements were developed and checked with 6 fire tests and 65 tests at room temperature [1, 2].

A simple calculation method providing a practical design rule to calculate the fire resistance of a usual type of composite steel and concrete beams is proposed. This method developed on the basis of experimental investigation permits to calculate, on the one hand, the load bearing capacity of continuous composite steel and concrete beams related to global element collapse through sagging and hogging moment resistances for a given fire duration. On the other hand, it takes account of local collapses of continuous composite beams under fire situation, such as the local resistance at supports related to steel local buckling and crushing. Furthermore, the method is also applicable to continuous composite beams with partial shear connection. Examples comparing the proposed simple calculation method and several fire tests will be shown in this paper, which gives evidence of the validity of the method.

Finite element method is applied for the calculations of temperature fields of concrete filled steel tubes under fire. A theoretical model that calculates deformations and strength of beam-column in fire and fire resistance is described in this paper. A comparison of results calculated using this model with the results of tests, there is good agreement. Based on the theoretical model, influence of the changing strength of the materials, sectional dimensions, steel ratio, load eccentricity and slenderness ratio on the fire resistance is discussed. The theoretical model in this paper were used to provide data on the necessary fire protection measures for the concrete filled steel tubular columns used in a high-rise building, SEG Plaza in Shen-Zheng City, South of China.

During the last decade, numerical models are being used more and more to analyse the behaviour of structures under fire action. In view of the limited dimensions of the test furnaces, the validation of these numerical models is mainly based on fire tests on single structural components. It is common practice, at least in Europe, to exclude structural restraint in fire tests. Finally, the fire is schematised as the socalled "standard fire", to which all isolated structural elements are considered to be exposed to. The reason is that structural restraint is hard to accomplish and differs from case to case. In real structures, there will often be -to some extent- structural restraint. As a consequence, the single-element standard fire approach is conservative. The ECSC co-sponsored project "The behaviour of a multi-storey steel-framed building subject to fire", offers unique possibilities to validate numerical models on the basis of realistic full scale fire tests on complete structural systems, including restraint effects. These "natural" fire tests are carried out at the Large Building Test Facility at Cardington. The potential of the FEM package DIANA to describe the thermal and mechanical response of fire exposed structural elements and systems, is illustrated on the basis of numerical analyses of natural fire tests on composite steel/concrete structural systems. It is demonstrated that structural restraint may significantly increase the fire resistance. Furthermore, it is demonstrated that significant savings can be made on passive fire protection measures. Finally the outline and the preliminary results of a parametric study are presented.

Fire safety is generally assessed on basis of required resistance time to a standard fire measured in classes with 30 minutes intervals. For structures dominated by bending failure, this approach seems appropriate to obtain a adequate safety level. However, this paper demonstrates that for other failure modes, performance criteria should preferably be expressed in target reliability levels rather than in time levels, as requirements in the time domain do not necessarily result in satisfactory safety. This has been illustrated by a characteristic comparison between a structure subjected to bending with one subjected to shear. For this purpose, a simple time dependent probabilistic simulation was carried using three consecutive calculation models: For this simulation, a method was developed to account for the accuracy of the used models.

The European Standards Organisation (CEN) has planned to develop a complete set of harmonized European building standards. This set includes standards for composite steel and concrete buildings and bridges. The Eurocodes, being the design standards, form part of this total system of European standards, together with standards for fabrication and erection and product standards. After a period of experimental use of the ENV (European Pre Standard)-versions of the Eurocodes, a start has been made with the conversion to EN's (European Standards). At the time of the conference, the 2^nd draft EN version of Part 1.1 of Eurocode 4: `Composite Structures — General Rules and Rules for Buildings' will have been completed and presented to CEN/TC250/SC4. An overview will be given of the conversion procedure, the structure of the Eurocode package and the planning for completion. Technical modifications in Part 1.1 of Eurocode 4 will be discussed in detail.

The Australian Standard AS 2327.1-1996 Composite Structures — Part 1: Simply Supported Beams has a number of aspects that differ from other existing codes and standards such as Eurocode 4. These include: construction stages and minimum construction loads; rational procedures for moment-shear interaction in conjunction with partial shear connection; simplified minimum shear connection requirements; reinforcement against rib shear failures; detailing requirements for shear connection using Australian steel sheet profiles to ensure ductile performance of the connection; and shear connection to tubes in composite beams.

A new composite column designed specifically to support large gravity loads in high-rise steel buildings is described. The column takes advantage of the in-plant prefabrication of steel, the economical compressive load carrying capacity of concrete, the speed of erection for a steel structure and the fire resistance inherent in a concrete column. The column is specifically designed to simplify the installation and removal of the form work needed to pour the concrete and requires less crane capacity to erect the column due to a much lighter steel section. The extensive test program carried out at three North American universities is summarised. The scope of application for the new column and the design equations that have been proposed are discussed. Finally, an example of the savings that can be achieved by using this column system is demonstrated by comparing the system to the actual design of an existing building. The new composite column has been patented in Canada, the United States of America and has received an International patent.

Steel-concrete composite columns may be designed either by requirements of the American Concrete Institute Building Code ACI 318-99 or by the American Institute of Steel Construction Specifications for Load and Resistance Factor Design, 2^nd Edition (1995). Each design standard is described for application to a concrete filled steel tube and to a concrete encased structural shape as each is designed for the same dimensional and service load conditions. These "standard" type column sections are used for the comparison, as the LRFD specification can be used directly only for such standard sections. The design exercise demonstrates that a) the LRFD specification requires fewer computational steps and is therefore easier to apply, b) the ACI rules tend to exaggerate the influence of slenderness, and c) different but very similar results were obtained for the two methods applied to the same design problem

The purpose of this research is to clarify the scale effect on the uni-axial compressive behavior of concrete-filled steel tube (CFT) short columns. Three different sizes of CFT short column specimens were tested. Other experimental parameters were the shape of the cross section, the strength of filled concrete and loading methods. Scale effects on the compressive strength of square CFT columns were observed, while those of circular CFT columns were not clearly observed.

Composite structures made of structural steel and concrete maximize the advantages of the two components. Using high-performance materials composite construction is becoming more popular. The European Coal and Steel Community has therefore supported the research project `Use of High Strength Steel S460' (1). Within the scope of this project, European research institutes worked on developing design and execution guidelines to overcome the obstacles associated with using these high-strength materials. The Institute for Structural Concrete has carried out experiments and performed FE calculations for composite beams made of High Performance Steel S460 (HPS) and High Performance Concrete C 70/85 (HPC).

This contribution describes the evaluation of recent experimental and numerical investigations performed in the frame of an European research project on composite structures with high strength materials [1]. The aim of the project was to determine the limits of applicability of existing codes and to improve existing design rules with regard to the use of high strength steels and high strength concrete. A testing programme was executed to provide information on the stability criteria and plastic design criteria for composite beams and columns and included the following experiments: 1) Tests on composite beams made of high strength steel in combination with normal strength or high strength concrete. 2) Tests on composite columns made of high strength steel in combination with normal strength or high strength concrete. 3) Static and dynamic push-out tests. The experimental results have been verified by numerical simulations that allowed to perform parameter variations and to generate results not fully covered by the tests.

This paper describes one solution to the vibration problem of long span composite beams by means of integrated dynamic absorbers. These were used in the floor construction of the new exhibition halls in Berlin. The stiff and heavy composite beams used in the original design with a natural frequency of 5 Hz were replaced by composite trusses with reduced weight and a natural frequency of less than 4 Hz together with integrated damper systems. Part of the bottom chord of the truss is a tension bar only fixed at the ends and vertically suspended by elastomeric springs. The tension bar stiffens the system and works additionally as a damper [1]. This method of construction means the weight of the beams can be reduced to 60 % of the comparable beam without dampers — a significant reduction in costs. The effects have been verified in tests. In addition, potentially interesting structures with integrated damper systems, used for instance in sports halls, are also described.

Design codes for composite structures include provisions for prototype load-testing, using small sample sizes. The paper compares alternative methods for treating load-test results and examines the influence of the type of sample statistic selected as the basis for estimating design strengths. It is shown that the mean strength is the most dependable sample statistic for the derivation of factored design strengths based on prototype testing with small sample sizes.

The clear advantages of launched solutions in the construction of composite bridges is analysed in two innovative construction solutions in order to increase their use in large span bridges and to open the way for possible lines of action for the greater employment of these techniques in the future.

A new strength model is proposed for composite beam-column joints in RCS frames consisting of steel beams and reinforced concrete columns. Based on connection subassemblage test data, the proposed model builds upon guidelines previously developed by the ASCE Committee on Composite Construction. The proposed model employs modifications to address unique failure modes and stress transfer mechanisms in the joints, and extends the scope of the earlier ASCE Guidelines to address a broader set of connection details. Comparisons between test data and calculated strengths demonstrate the accuracy of the proposed model.

The seismic behavior and design of hybrid coupled shear walls is discussed. A prototype system is designed according to current specifications then analyzed using the finite element approach according to criteria in FEMA-273 (1). The performance of the system and its individual components is evaluated and seismic design provisions are assessed.

The focus of this paper is on cyclic response of "shear" connections between steel outrigger beams and reinforced concrete core walls. The experimental program was conducted in two phases. The paper summarizes the first phase during which seven specimens with a number of variables were tested. The connections involved a beam bolted to a shear tab that was welded onto a plate embedded inside the wall. Multiple anchors were used to connect the stud plate to the wall. Test results show that the actual capacity is larger than the design values; however, the mode of failure was brittle (either through pull-out of the embedded plate or by weld fracture at the stud-plate interface) which has to be avoided. A model was developed for accurate prediction of the capacity, and was verified based on the experimental data. This model was used in a capacity design approach in order to dissipate energy in a ductile manner through yielding of the shear tab. The second phase of the research was intended to further examine the cyclic behavior of outrigger beam-wall connections under more realistic loading and boundary conditions. Preliminary observations from this phase are also summarized herein.

This research investigates the seismic behavior of partially-restrained steel frames with composite reinforced concrete infill wall panels. Headed shear stud connectors are used to attach the infill panel to the steel frame to achieve composite interaction. The research program includes twelve experiments of the combined shear and tensile response of shear studs in infill wall panels, two one-bay, two-story, one-third scale experiments of the complete structural system, and development of a nonlinear cyclic finite element formulation for simulating the seismic response of the system. This paper summarizes the results of this ongoing research, focusing primarily on the findings to date of the experimental research.